DeepSeek
DeepSeek R1
Open-weight reasoning model that holds its own against o3-mini.
ReasoningValue
DeepSeek R1 is not currently offered on Token Harbor.
Looking for something comparable? DeepSeek V4 Flash is live on our catalog — same family, currently served and priced.
Browse live models →Ratings & benchmarks
Snapshot 2026-04-28Overall
4.4 / 5
Composite of intelligence + reliability.
Output speed
40 t/s
Output tokens per second under typical load.
Time to first token
8.0 s
Lower is better. Reasoning models naturally stretch this.
Value
4.4 / 5
Intelligence per dollar at typical mix.
Context window
163,840 tokens